1. Reliabiity refers to the stability if a test score over repeated administrations. A reliable test will yield stable scores over repeated administration, assuming the trait being measured has not changed.
1.1. There are four types of estimates of reliability: Test ReTest estimates of reliabilty, Alternative form estimates of reliability, Internal consistency estimates of reliability, and Internal consistency estimates of reliability.
1.1.1. Test Retest esitmates of reliability are obtained by administering the same test twice to the same group pf individuals, with a small tine interval between testing, and correlating the scores.
1.1.2. Alternative form estimates of reliabilty are obtained by adminitering two alternate or equivalent forms of a test to the same group and correlating their scores. The time interval between testings is short as possible
1.1.3. Internal consistency estimates of reliability fall into to catergories: spilt half or odd even estimates and item total correlations
1.1.3.1. Split half and Odd Even estimates divide a test into two halves and correlate the halves with one another
1.1.3.1.1. These correlations are based on half test the Spearman Brown prophecy formula is used to correct these estimates to what they would be if based on a whole test
1.1.4. Internal consistency estimates tend to yeild inflated estimates for speeded test.
2. A valid test measures what it is suppose to measure.
2.1. There are three types or validity evidence: Content evidence, Criterion-related validity evidence, and Construct validity evidence
2.1.1. Content Validity Evidence: is assessed by systemically comparing a test item with instructional objectives to see if they match.
2.1.2. Criterion Related Validity: is established by correlating test scores with an external standard or criterion to obtain a numercial estimate of validity evidence.
2.1.2.1. There are two types of criterion related validity evidence; Concurrent and Predictive Validity Evidence
2.1.2.1.1. Concurrent Validity Evidence is determinded by correlating test scores with criterion measure collected at the same time.
2.1.2.1.2. Predictive Validity Evidence is determined by correlating test scores with a criterion measure collected after a period of time has passed
2.1.3. Construct Validity Evidence is determined by finding whether test results correspond with scores on other variables as predicted by some rationale or theory.